April 1990

in association with the Australian Academy of Science. The then BMR Division of Petrology & Geochemistry played a major coordinating role. Contributions from 59 authors have been com•bined into a single comprehensive account by the editors: R.W. Johnson, 1. Knutson (both BMR), and S.R. Taylor (Research School of Earth Sciences, Australian National University, Canberra).

The volume brings together the vast amount of volcanological, petrological, geochemical, geoch•ronological, geophysical, and tectonic data now available relating to east Australian and New Zea•land intraplate volcanism. As such, it con tributes substantia lly to the understanding of the geologi•ca l evolution of eastern Australia and New Zea•land in particular, and also of intraplate volcanism generally and its role in crusta l and continental development.

Quasi-continuous sampling of the mantle

Intermittent vo lcanism has taken place a long the 4400 km- Iength of eastern Australia from about 70 Ma to 4000 B.P., and represents an except ional example of long-term, widespread volcanic act ivity (Fig. 15). The abundance of mantle xenoliths and megacrysts in the basaltic rocks is also unusual, and represents one of the largest and most widespread quasi-continuous samplings of the mantle section on the Earth's surface, both in space and time.

The Cainozoic volcanic rocks of eastern Aus•tralia are characterised by: (I) a dominance of maric magmas of diverse compositions, (2) the importance of fractional crystallisation in the deve lopment of the magmas, and (3) relatively diverse geochemical and isotopic composit ions throughout the range of magma types, suggesting possible differences in magma source and evolution.

The distribution of the east-Austra li an volcanic provinces on the uplifted flank of a rifted continen•tal margin strongly suggests contro l relating to the tectonics of the continental margin.

Two models are suggested in the volume for the origin of the east-Australi an volcanism. In the first, uplift of the eastern highlands is attributed to a combination of the thermal anomaly generated during continental extension and the underplating of mantle-derived igneous rocks during lithos•pheric extension. This model predicts a regional thermal peak between 80 and 50 Ma that has been decaying ever since, except where locally per•turbed by hotspots that have triggered relatively local volcanic and subvolcanic activity. A second interpretation is that the volcanism has been con•trolled large ly by intraplate stresses in the Indo•Australian plate, caused by its interaction with surrounding plates. Much of the volcanism has taken place a long the relatively young Tasman Fold Belt System, which, being adjacent to a man•tle anomaly that produced late Mesozoic and early Tertiary seafloor spreading, is underlain by hotter mantle than in cratonic areas.

Sapphire, corundum, zircon, and pleonaste arc present in alluvial and weathered deposits in many of the volcanic lields in eastern Australia, and are becoming of increasing economic importance. Diamond has a lso been reported in alluvial set•ti ngs at numerous localities. Early work has indi•cated that the high geothermal gradient in eastern Australia makes any direct association between Cainozoic volcanism and diamond unlikely, and a

Fig. 15 (opposite). Intraplate volcanic provinces of eastern Australia (adapted in part from O'Reilly & Griffin, in Nixon, P.". (Editor), 1987: MANTLE XENOLITHS. Wiley, Chichester).

derivation from Mesozoic pipes has been sug•gested from nitrogen-aggregation studies of Cope•ton diamonds (p.12, this issue).

Do lamprophyres have high precious•

metal contents? A number of gold deposits are spatially and, in some cases, genetically associated with alkaline rocks, especially felsic stocks and porphyries of calc-alkaline to alkaline composition (e.g. Mutschler & others, 1985: Transactions of the Geological Society of South Africa, 88, 355-377). This association has been highlighted by a recent proposal that lamprophyres in general have inherently high levels of Au (Rock & Groves, 1988: Nature, 332, 253-255) and platinum-group elements (Rock & others, 1988: in Geology Department & University Extension, University of Western Australia, Publication 12,295-308). As part of its program on alkaline igneous rocks and related mineralisation, BMR has obtained precious-metal abundances on a range of alkaline rocks with a view to establishing their background precious-metal inventory.

New data on the Ellendale and Argyle lampro•ites, severa l Australian carbonatites, kimberlites from South Africa and India, picritic monchiquites from the Carnarvon Basin, and a lkaline volcanics from the Stuart Shelf indicate much lower levels of Pd. PI. and Au « 10 ppb, commonly ~. 5 ppb each) than has been attributed previously to such rocks. Post-mineralisat ion lamprophyres from Tennant Creek have low to moderate contents of Au and PGE. Altered K-rich volcanic rocks from the Redbank area, NT, and metamorphosed lam •prophyres from the Leonora area of the Yilgarn Block generally have low abundances ofPGE and Au although a number have anomalously high PGE <Redbank) and Au (Yilgarn) abundances. The wide range in abundances and the distinctly non-chondri tic abundance ratios observed in these anomalous samples are believed to result from mobilisation, particularly of Au, during low•temperature metamorphism and /or hydrothermal alteration.

Modern analytical techniques Published precious-metal values in alkaline

rocks show a very wide range and are confused by the inclusion of mineralised samples and analyses by inferior techniques. Modern analytical tech•niques such as neutron activation and inductively coupled plasma mass spectrometry (tCPMS) with low limits of detection (e.g .~ 0.5 ppb) have over•come many of the problems In assessing the intrin•sic or background precious-metal contents. in unmineralised rocks, which requires accurate determination at the < 5 ppb level. Problems remain. however, in terms of heterogeneous dis•persion of metals and contamination. as well as the geological problems such as non-representative sampling and the effects oflow-temperature alter•ation on primary values. The Pd. Pt, and Au ana•lyses reported here were performed using ICPMS on the least-altered material available, following Pb collection by tire assay. Full PGE analyses were obtained by NiS collection and ICPMS for a limited number of samples.

Fig. 16. Plots showing mean and I-sigma varia•tion (bars) of metal and sulphur abundances in a range of alkaline rocks normalised to the mantle values suggested by Barnes & others (1988: in Geo-Platinum. 87,113-143).

BMR Research Newsletter 12

For further information contact Dr Wally Johnson or Dr Jan Knutson at BMR (Minerals & Land Use Program).

10r--------------------------. LAMPROtTES

Olivine Lamproite Leucite Lamproite Argyle

O. I f-...L........1..---' __ -'--'----' __ '--...L...--...L---'--I

10

0.1

CAR BONATtTES

, ~

I \ / \, , / '\ ,,/ \

'i \ ,

Redbank Mud Tank Cummins Range MtWeld

0.01 '--:==========~ 100 c-

10

0.1

LAMPROPHYRES

Tennant Creek Yilgarn ILeonora)

0.01 '-...L--...L--.JL-..L....--'----' __ L-...L.....J..--.JL...J Ni Os Ir Ru Rh Pt Pd Au Cu S

18/ E52- 2/ 10

17

BMR Research Newsletter 12

Kimberlites and lamproites Previous analyses of Au and PGE in South Afri•

can and Indian kimberlites by modern methods indicate a wide range of abundances: 1- 19 ppb Pd (average 8 ppb), 0.5 - 6 ppb Ir (average 3 ppb) and 0.1 - 43 ppb Au (average 12 ppb: Paul & others, 1979: in KIMBERLITES. DIATREMES. & DIAMONDS: THEIR GEOLOGY. PETROLOGY. & GEOCHEMISTRY. Blackwells Scienlific Publications, Melbourne, 272- 279).

Rock & others (1988) suggested that lamproites may also have high intrinsic Au and PGE levels. However, new data obtained for Argyle and West Kimberley lamproites and several kimberlites from India and southern Africa do not support this contention. Low values of PGE and Au were obtained for both the kimberlites and the West Kimberley leucite lamproites: - 2 ppb Pd, - 2 ppb Pt, and 1- 2 ppb Au, and only slightly higher values were obtained for olivine lamproites (> 20% MgO) from Ellendale: - 3 ppb Pd, 4 ppb Pt, and - 2 ppb Au (Fig. 16a). These values cannot be taken to indicate a mantle source enriched in PGE and Au. From a study of the PGE content of mafic - ultra•mafic rocks in the Kimberley and Pilbara regions, Sun & others (Precambrian Research, submitted) concluded that fertile mafic-ultramafic magmas commonly have - 15 ppb Pd and Pt, and that the convecting mantle has had fairly homogeneous PG E abundances (- 4 ppb Pd, - 6 ppb Pt, - I ppb Au) throughout geological time.

Lamprophyres Higher precious-metal contents - 7 ppb Pd, 6

ppb Pt, and 4 ppb Au - were found in Jurassic lamprophyres (picritic monchiquites) from the Carnarvon Basin which have also been prospected for diamonds (Jaques & others, 1986: GSWA Bul•letin 132). These values are considerably higher than those reported by Mitchell & Keays (1981: Geochimica et Cosmochimica Acta, 45, 2425 - 2442) for Victorian basanites <0.7 - 1.7 ppb Pd, - 0.3 ppb Au).

Lamprophyres - mostly minettes - are com•mon in the Tennant Creek area and many have high MgO( 10-20%), Ni andCr(> 300 ppm Ni, > 400 ppm Cr), implying a mantle origin. The lam•prophyres have a wide range in Au and PGE contents and include some with up to - 30 ppb Pd, 30 ppb Pt, and 10 ppb Au (Fig. 16c). PGE patterns are broadly chondritic in form, suggesting that the PGE and Au abundances are primary, though it is possible that the precious metals have been en•riched because the lamprophyres intrude minera•lised rocks (Black, 1977: BMR Journal of A ustral•ian Geology & Geophysics, 2, 111 - 122) that have high PGE contents as well as Au.

Lamprophyres are also widespread in the gold•fields region of the Yilgarn, and their spatial asso•ciation with some Au deposits has led to specula•tion (Rock & Groves, 1987: Geology, 16,538- 541) that fluid-rich lamprophyres may be the source of Au in Archaean mesothermal gold deposits. New data obtained by BMR during I: I 00 000 scale regional mapping in the Leonora region indicate thai many of the lamprophyres do not have high Au (or PGE) contents. Samples that do have anomalously high Au have PGE abundances which depart markedly from chondritic abun•dance (Fig. 16), suggesting that the Au is intro•duced rather than primary. The Au may have been introduced by metamorphic fluids because many of the lamprophyres have been metasomatised and /or metamorphosed. The low MgO, Ni , and Cr contents of most of the lamprophyres are incon•sistent with mantle derivation , and high Au or PGE levels in these rocks cannot be taken to indicate Au- and /or PGE-rich mantle sources. Lamprophyres spatially associated with gold min•eralisation in the Superior Province of Canada have also been shown to have low Au contents excepl where mineralised (Wyman & Kerrich, 19X8: c('onomic Geology, 83,454- 461 ).

Platinum-group metals are recovered from min•ing of the Precambrian porphyry-style copper deposit hosted by the Phalaborwa carbonatite

April 1990

complex in South Africa. However, data obtained on the Mount Weld, Mud Tank. and Cummins Range carbonatites in Australia indicate compara•tively low levels of precious metals, averaging 2 ppb Pd, 2 ppb Pt, and 2- 3 ppb Au with individual samples up to 3 ppb Pd and Pt, and 7 ppb Au (Fig. 16b). The high abundances of PGE and Au in the Phalaborwa Complex, therefore, do not appear typical of carbonatites. Samples from the Cu•bearing breccia pipes at Redbank , which are believed to have affinities with carbonatites (Knutson & others, 1979: Economic Geology, 74, 814- 826), commonly show higher values ranging up 30 ppb PI: a carbonate-rich mineralised (pyrite•bearing) breccia gave 330 ppb Pd, 22 ppb Pt, and 120 ppb Au.

Exploration models The new data call into question previous specu•

lation concerning intrinsic Au and PGE levels in lamproites, kimberlites, and a wide range of lam•prophyres, and their possible role in the formation of Au and PGE deposits. Reliable published Au and PG E data on these rocks are scant and the new data highlight a need for better characterisation of background levels of precious metals in alkaline rocks generally. using modern precise analytical methods. Precious-metal enrichment in Austral•ian alkaline rocks documented here is accompan•ied by pervasive alteration, notably K-metasom•atism. oxidation, and carbonation (CO?). These alteration facies, together with phyllic aiteration. are characteristic of precious-metal deposits associated with alkaline and calc-alkaline rocks in general rather than a particular rock type. Never•theless, certain calc-alkaline to alkaline suites, such as high-K calc-alkaline rocks and shosho•nites (Wyborn & Cameron, 1990: Australian Geo•logical Convention, Hobart, Abstracts), appear to have inherent high background values ofPGE and Au which make them pOlentially attractive explo•ration targets.

For/imher injimnation contact Dr Lymon Jaques at BMR (Minerall' & Land Use Program).

The Giles Complex, central Australia New insights into tectonics and metamorphism

The western Musgrave Block (where Western Australia meets the border between South Aus•tralia and the Northern Territory) contains the largest exposures of deep-seated basic-ultra basic intrusions in Australia, and thus has special sig•nificance for the understanding of crustal struc•ture and of potential mineralisation in central Australia. Recent investigations of the layered basic-ultra basic Giles Complex (Proterozoic) in the Tomkinson, Blackstone, Murray, Cavanagh, and .Jameson Ranges in Western Australia by BMR during 1987 and 1988, in collaboration with the University of Tasmania, the University of Melbourne, and the Geological Survey of Western Australia, have resulted in new interpretations of the structural and metamorphic evolution of this high-grade metamorphic terrane.

The layered intrusions of the Giles Complex occur mainly as faulted segments in granulite•facies terrane. This terrane is located to the south (and constitutes the hanging wall) of the Wood•rolfe Thrust. a major east - west lineament where northward thrusting of granulite-facies rocks over amphibolite-facies rocks south of the Amadeus Basin (Collerson & others. 1972: Journal o{the Geological Society oj"Australia, 18,379- 394) mir•rors the southward thrusting along the Redbank lineament north of the Amadeus Basin (Glikson. 19X6: Transactions o{the Geological Society o{ South Ajrica. 89, 263- 2X3: Glikson. 19X7: BMR Journal. 10, X9- 1 07). The outcropping of the deep

18

crustal layered intrusions of the Giles Complex and of equivalents at Mount Hay. north of the Redbank lineament. is significant for understand•ing the deep crustal structure in central Australia as interpreted following BMR's recent seismic rellection transects in the region (Goleby & others. BMR Research Newsleller, 7,8- 9: Goleby & others. 19X8: Nature, 337, 325- 330).

The Giles Complex is well exposed in the Tom•kinson Ranges, SA and W A (Fig.17), where ear•lier studies by Thomson (1975: Australasianlnsti•tute or Mining & Metallurgv. Monograph 5, 451 - 460) and Danicls( 1974: Geological Survevo{ Westert/ Australia, Bulletin 123) outlined the regional framework and where detailed tield and petrological studies have been conducted on some of the layered intrusions, including Kalka. Ewar•ara. and Gosse Pile (Nesbitt & others. 1970: Geo•logical Society of South Africa, Special Publication 1,547 - 564). These studies and recent BMR work in Western Australia (BMR Research Newsleller. 10,3-4: Ballhaus & Glikson. in press. Journal oj" Petrologv) have allowed the major struclural cle•ments of the western Musgrave Block 10 be idcnlilied.

Structural and stratigraphic outline The layered basic- ultrabasic units occur as sli •

ces separated and cut by mylonitic shear zones and intruded by moslly porphyritic (rapakivi) graniles.

Aeromagnetic maps produced by BMR allow delineation of major structural discontinuities in the Musgrave Block. The Woodrolle Thrusl stands out as a boundary between a magnetically dis•turbed terrane to the south (granulite and amphib•olite facies rocks) and a magnetically quiet terrane to the north (amphibolite facies rocks). Major faults, such as the Mann, Davenport, and Hinckley Faults, are marked in part by linear anomalies. Tectonic slices of the Giles Complex in the Tom•kinson Ranges are distinct from intervening gran•ites and felsic granulities, while the recrystallised basic granulites of the Hinckley intrusion (Fig. 17) are expressed as a magnetically quiet zone com•pared to the gabbro. The Bell Rock intrusion is bounded on the west by a northwest-trending line•ament interpreted as a major fault. This fault separates the more strongly deformed sector of the Giles Complex (including the Bell Rock. Hinckley. Michael Hills. Mount Davies. and Kalka layered bodies) from the shallow-dipping. least deformed sector (including the Blackstone. Cavanagh. and Jameson layered bodies).

Earlier work and the present study allow delini •tion of several principal geological units. with the following timc sequence:

Banded felsic gneisses, showing composite mineralogical banding (S I) and early intrafol•ial I(llds that deform an early fabric and line•ation. are of probable volcanic origin (Gray,